Subharmonics in Vintage Ham Equipment
or, Why Pi Network Interstage Coupling Doesn't Work
  

Subharmonics - a topic which has largely disappeared from ham literature in recent years, was a serious problem in equipment designed in the 1950s and 60s. I am presenting it here as a caution to newcomers to restoring and operating vintage equipment. It came up for me in the process of rehabilitating a Hallicrafters HT-40, which is the subject of another article on this web page.

What is a subharmonic? It is a signal that passes through into the antenna from a transmitter. It is an undesired signal that is calculated by dividing the desired signal by an integer. For example, if you wish to transmit on 3.885 MC, there is also a 1.9425 MC signal present in the transmitter output. The transmitter does not know how to divide, a digital technique it does not possess. The fundamental VFO frequency of 1.9425 MC is passed through the stage that is intended as a doubler, due to inadequate internal filtering.

The early transmitters used link coupled parallel tuned circuits to couple to the antenna. Most also used parallel tuned circuits in the multiplier and buffer stages. The VFO typically had two ranges, one operating on 1.75 to 2 MC, and the other operating on 7.0 to 7.4 MC. The Collins 32V operated only in the 2 MC range for everything, and had traps in the exciter stages to prevent unwanted harmonics of the 2 MC PTO from entering the final amplifier. The low VFO frequency allowed exceptional stability and ground breaking dial accuracy. Cost dictated the choice of fewer VFO frequencies and minimum oscillator coil and capacitor combinations in less pricey rigs.

In the 50s and 60s, the Pi Network coupling circuit replaced the tricky to adjust and expensive link coupling networks. The Pi Network also lent itself well to faster to operate band switching. Opening a transmitter cabinet to plug in different coils for band switching prevented TVI reduction techniques and exposed the operator to shock hazards. The Pi Network in the antenna coupler circuit offered superior TVI reduction, since it was a low pass filter.

Designers of the Apache, DX100 and other popular transmitters adopted the Pi Network for interstage coupling, and that is where the problem of subharmonics began to appear. As a LOW PASS filter, it allowed the fundamental VFO frequency to couple through to the final amplifier in a multiplier stage. The Apache and DX100 corrected for this by including a parallel tuned circuit in the grid input circuit of the last straight through buffer stage. The Johnson Viking 2 and Valiant, and even the economy DX60 and some versions of the Globe Scout used parallel tuned circuits all the way through. You should examine the schematics of a vintage rig that you are considering acquiring to see if parallel tuned circuits are employed in the drivers. Be suspicious of any rig that uses pi network interstage coupling.

Pi Network interstage coupling was highly touted as suppressing VHF parasitics by virtue of the capacitor connected from grid to ground that is part of the Pi Network that matches the high plate impedance of the driver tube to the low impedance of the class C final. This did not work out so well, as evidenced by neutralization fixes for nearly all of the transmitters in that period. The later versions of the HT-40 and Apache came from the factory with wires near the plate of the final tubes that provided some neutralization. Even the DX100 had a fix that is available on the net today.

To add a parallel tuned circuit to the grid of the last buffer/driver stage in the HT-40 would require an extra band switch wafer and the associated inductance and capacitance. This would preserve the original design without changing to a high impedance parallel tuned circuit in the grid circuit of the final amplifier. A parallel tuned grid circuit in the final would make neutralization mandatory. Retrofitting such an arrangement would be a major project.

Some tubes, such as the 6L6, are not well screened, and can be operated this way, but the adjustment can be tricky. That is why you rarely see the 6L6 used above 40 meters. A successful 6L6 rig appears in the 1959 Radio Amateur's Handbook on page 184. It claims 75 watts input on up to 15 meters. I have considered building this one with a matching push pull class AB2 6L6 modulator, just to have the experience of doing it. There was a 7868 tube with a brief period of popularity which was a smaller 6L6. It had short internal leads and found some use as a multiplier stage in VHF transmitters. It might work well in this 1959 rig up to 10 meters, if another tube is added to the exciter. The advent of cheap sweep tubes, the surplus 807s, and the new 6146 spelled the end of the 6L6 tube's dominance of the low end ham market. Some of the sweep tubes of the day were a little squirrely on higher frequencies, but even the popular Yaesu FT101 series used sweep tubes for years. There are articles out on the web, even a QST article mod to convert the FT101 to 6146 operation. Now that TV sweep tubes are not so cheap, this is an option.

The other fix to use a rig that uses Pi Network interstage coupling is to build an external parallel tuned buffer for the Heathkit VF-1 or Johnson 122 to clean up the signal before putting it into a rig like the HT-40. You can also use a Hallicrafters HA-5 VFO. But they are relatively rare and expensive. They are very stable by comparison to other equipment of the day, since they are heterodyne VFOs ahead of their time. The easiest choice that uses non vintage parts is a modern solid state synthesized VFO like the N3ZI unit, and operate it on the frequency of transmit output, so that there is no possibility of other frequencies coupling to the antenna, since everything operates “straight through”.

When the subharmonic problem manifests, the rig appears to tune up OK. The only indication of problems is high SWR if you are using a mono band antenna. The SWR is higher than anticipated because the unwanted spurious output does not match well into the antenna tuned to the desired frequency. If you check the output RF with a monitor scope as you tune the PLATE and LOADING controls, you may clearly see the lower frequency RF. A spectrum analyzer, if you have it, will clearly show the problem. A friend a few miles away can listen on a receiver. Whenever you put a vintage rig on the air, check it carefully.

An antenna coupler like the Johnson Matchbox is parallel tuned circuit that is a good way to reduce any low level spurious products in the transmitter output. I also use a T match low pass circuit described elsewhere that suppresses harmonic radiation of vintage rigs into multi band antennas. However, it would not help with subharmonics, since it is a low pass configuration.

I hope this article clarifies the problems and some of the solutions for vintage rigs that have subharmonic output problems.

  

73,
Janis
AB2RA
Text size:  +   –      Back to Vintage Equipment
Back to AM Transmitters
Back to the Home Page